Re: OS/400 command set

No, HTTP is not "connectionless".   When you get something from a web
server, you connect, you request a document, it sends it, you request
the next, it sends it, etc... all during one TCP connection.

What you are thinking of is that it's "stateless".  Each file that you
request is it's own request, unrelated to all others.  This mainly makes
sense when working with CGI scripts or similar "program-driven" requests.
You send variables to the web server, and it sends back data.   That's
considered the entire life of the transaction.   When you need to request
the "next screen" of information, it's a whole new "unrelated" request.
The server calls teh CGI program again, as a new program call.  You can
use a variable embedded in the web page to try to keep track of where you
left off, but all of the data has to be saved from request to request,
since (from the HTTP server's perspective) each request is unrelated.

The reason for having multiple simultaneous connections is to download
the main HTML, plus pictures, sounds, alternate frames, etc all at the
same time.   Each picture, sound, frame, applet, etc on a web page is
a whole new document... they're just displayed together to make the
web page look nice.   So, web browser usually download the various
elements simultaneously to make things load efficently, etc.

On Thu, 14 Nov 2002, Leif Svalgaard wrote:
>
> just to educate an ignoramus:  I thought web serving was basically
> connection-less (each transaction stands on its own with no context
> or memory about other transactions - except what might be carried
> around within the HTML text itself), so what does "simultaneous
> connections" mean?
>